I love Netflix. It streams all season of How I Met Your Mother and all but the latest season of Mad Men. What more could I ask for? Apparently a lot. With my limited down time, I mostly stick to my favorite shows, and for the rare occasions I do deviate from my TV-watching-routine, I usually settle for one of the TV shows Netflix recommends for me. It’s takes the guessing of what I’ll like out of the picture.
After reading AdAge’s article, “Data-Mining Boosts Netflix's Subscriber Base, Showbiz Clout” I found out that 75% of the TV shows people watch on Netflix are introduced to them through the “Recommendations” driven by a data-mining program.
Here is an excerpt from AdAge’s article on Netflix’s power harnessing for recommendations and churn management.
“Even more impressive than the way Netflix is attracting new subscribers is how it uses data to retain users during a time when cable and satellite operators are bleeding customers. Netflix no longer discloses churn, but Janney Capital Markets analyst Tony Wible estimates it's about 3.5% to 4% per month. With former Warner Bros. exec Kelly Bennett assuming the CMO post in July 2012, the company is using data to serve up the right content and marketing to the right person. Netflix tracks what its users watch, search for and rate, as well as the time of day, date and device used. It even follows how subscribers browse and scroll. This allows it to know you're more likely to watch a romantic comedy Saturday night and stream from your tablet Monday morning. Netflix has said 75% of viewer activity is driven by recommendations.
Using its data insights, the company has moved away from boasting a vast library of movies that can be found on other services and is instead focusing on exclusives and specific genres like TV.”
How’d they do that?
When I dug a little bit deeper and found the article by The Motley Fool “3 Takeaways You Might Have Missed When Netflix Reported Earnings” I discovered that the intricate web of data-mining that allows Netflix to do recommendations based on your preferences that took them close to 10 years to fine tune.
“More than 75% of the viewing hours on Netflix start with the recommendations system. That's a huge number, and it underscores the competitive advantage Netflix built over the last decade or so. In short, Netflix is pretty good at figuring out what you want to see.”
Your company can do the same!
What took Netflix 10 years to build, you can accomplish in the next 90 days: 75% of your revenues could be driven by recommendations - now that predictive recommendation algorithms are available in an easy to use cloud platform and you no longer need data scientists in house to pull it off.
AgilOne brings you top-notch data science and results like those of Netflix without the headache of hiring and managing predictive analytics infrastructure and staff.